Distributed modifier-adaptation schemes for the real-time optimisation of uncertain interconnected systems

ABSTRACT

This paper proposes a set of distributed real-time optimisation schemes for the steady-state optimisation of parametrically and structurally uncertain systems that are composed of multiple interconnected subsystems. These schemes are derived in the framework of modifier adaptation and use uncertain system models of varying complexity. Consequently, each distributed modifier-adaptation scheme imposes different requirements on the available plant measurements, the communication network, and the control architecture. We prove that every scheme can converge to the plant optimum despite model uncertainty and compare the main features of each approach. Finally, we illustrate our findings via a numerical example.

KEYWORDS:

• Distributed optimisation
• interconnected systems
• real-time optimisation
• modifier adaptation
• model uncertainty

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Swiss National Science Foundation [200021E-158644].